Abstract
Objective: To examine the functional consequences of episodic ataxia type 2 (EA2)-causing nonsense and missense mutations in vitro and to characterize the basis of fluctuating weakness in patients with E2A.
Background: Mutations in CACNA1A encoding the Cav2.1 calcium channel subunit cause EA2 through incompletely understood mechanisms. Although the Cav2.1 subunit is important for neurotransmission at the neuromuscular junction, weakness has not been considered a feature of EA2.
Methods: The disease-causing mutations in three unrelated patients with EA2 and fluctuating weakness were identified by mutation screening and sequencing. Mutant constructs harboring mutations R1281X, F1406C, R1549X were transfected into COS7 cells and expressed for patch clamp studies. Single-fiber electromyography (SFEMG) was performed in patients to examine synaptic transmission at the neuromuscular junction.
Results: Functional studies in COS7 cells of nonsense and missense EA2 mutants demonstrated markedly decreased current densities compared with wild type. SFEMG demonstrated jitter and blocking in these patients with EA2, compared with normal subjects and three patients with SCA-6.
Conclusion: EA2-causing missense and nonsense mutations in CACNA1A produced mutant channels with diminished whole cell calcium channel activity in vitro due to loss of function. Altered biophysical properties or reduced efficiency of plasma membrane targeting of mutant channels may contribute to abnormal neuromuscular transmission, manifesting as myasthenic syndrome.
Episodic ataxia type 2 (EA2) is a rare, dominantly inherited disorder characterized by attacks of ataxia lasting hours to days with interictal eye movement abnormalities. These attacks are often triggered by fatigue or emotional stress and can be dramatically responsive to acetazolamide, a carbonic anhydrase inhibitor. Different disease-causing mutations in the gene CACNA1A have been identified in patients with allelic disorders EA2, familial hemiplegic migraine (FHM), and SCA-6.1,2⇓ CACNA1A encodes the Cav2.1 (formerly known as the α1A) subunit, which is the pore-forming and voltage-sensing subunit of the P/Q-type calcium channel complex3,4⇓ expressed throughout the CNS but most abundant in the cerebellum.5 The Cav2.1 subunit is also present at axon terminals of motor neurons. Autoantibodies directed against the Cav2.1 subunit cause paraneoplastic Lambert-Eaton myasthenic syndrome, thus emphasizing its importance in synaptic transmission at the neuromuscular junction.6 Whether mutations in CACNA1A cause congenital myasthenic syndromes has not been demonstrated.
Voltage-gated calcium channels are multimeric membrane complexes that open in response to membrane depolarization to let Ca2+ into the cell, where Ca2+ serves as an important second messenger. The Cav2.1 subunit has four homologous domains (I through IV), each with six transmembrane regions (S1 through S6). S4 is an integral part of the voltage sensor; S5 and S6 as well as the P loop interconnecting S5 and S6 likely form the channel pore. The auxiliary β, α2δ, and γ subunits interact with the Cav2.1 subunit and are important in the assembly, membrane targeting, and kinetic modulation of the voltage-gated calcium channel complexes.7
How mutations in CACNA1A cause clinical symptoms and signs is not well understood. Previous studies have shown that missense mutations causing FHM and glutamine-encoding CAG repeat expansions causing SCA-6 altered the biophysical properties of the calcium channel complex.8-10⇓⇓ A recent report noted the complete loss of function of a missense EA2 mutant construct expressed in HEK293 cells.11 Little is known regarding the functional consequences of nonsense and splice site mutations causing EA2, which have been hypothesized to form truncated proteins that may or may not be targeted to the plasma membrane.1,12⇓
Three unrelated patients with acetazolamide-responsive EA2 and fluctuating weakness were examined. The clinical profile of two unrelated patients and corresponding disease-causing nonsense mutations in CACNA1A have previously been reported.13,14⇓ A novel missense mutation was identified in a third patient. We studied whether these nonsense and missense mutations causing EA2 altered the function of calcium channel complexes in vitro by expressing the mutant constructs in COS7 cells, which are immortalized monkey kidney epithelial cells that, unlike HEK293 cells, had no endogenous voltage-dependent calcium channel subunits.15 Furthermore, given the prominent complaint of weakness, we examined neuromuscular transmission by single-fiber electromyography (SFEMG).
Subjects and methods.
Cases.
Patients 1 and 2 have been described in detail.13,14⇓ In brief, since early childhood Patients 1 and 2 have experienced attacks of vertigo, truncal and limb ataxia, and slurred speech, often triggered by vigorous exercise or stress. They also had diffuse weakness during ataxic spells. Neurologic examination was notable for interictal gaze-evoked and rebound nystagmus and mild truncal ataxia.
Patient 3 is a 45-year-old man who first developed episodic weakness in his teens. Previously an avid surfer, the patient was unable to continue surfing because of attacks of weakness. He was thought to have MG, for which he underwent extensive evaluation. Results of nerve conduction studies and EMG were normal, and antiacetylcholine receptor antibody titers were negative. He reported improvement in his symptoms in response to pyridostigmine.
In his 30s, the patient began to experience violent spinning sensations and imbalance in addition to spells of weakness. He denied any history of headache. On examination, he had mild truncal ataxia with rebound and gaze-evoked nystagmus. In response to acetazolamide, he experienced a dramatic decrease in the frequency and severity of his attacks of dizziness, imbalance, and weakness.
He was unaware of similar symptoms in his father. His mother and two siblings were healthy with no symptoms or signs suggestive of migraine or EA2 (figure 1).
Figure 1. Pedigree of Patient 3 (filled symbol with arrow) with no other affected family members (open symbols). Haplotypes for markers on chromosome 19p are shown.
Genotyping.
In Patient 3 and his family, we isolated DNA from consenting subjects and typed a series of microsatellite markers on 19p flanking CACNA1A: D19S221, D19S1150 (intragenic), and D19S226.
Mutation screening and identification.
To screen for polymorphisms, we used single-strand conformation polymorphism (SSCP) and denaturing high-performance liquid chromatography (HPLC). Direct sequencing was performed to identify specific nucleotide changes.
Functional study of mutant channels in vitro.
We used the 1A-1 isoform (generously provided by SIBIA, La Jolla, CA; GenBank accession number AF004884) as our wild-type template, because this splice isoform, which contains the glutamine-encoding CAG repeats in its reading frame,16 may be preferentially expressed in neurons.17 The F1406C and two previously reported EA2-causing R1281X and R1549X13,14⇓ mutations (based on AF004884; corresponding to 1279, 1404, 1547 in GenBank X99897) were introduced by a four-primer mutagenesis method18 or QuikChange (Stratagene, La Jolla, CA) using paired forward/reverse mutagenesis primers. The mutant constructs were confirmed by sequencing.
Plasmids containing the mutant or wild-type 1A-1 cDNA and those encoding the α2δ and β4 subunits were cotransfected by lipofection (Lipofectamine2000, Life Technologies, Carlsbad, CA) into COS7 cells, along with a reporter plasmid for CD8, at a mass ratio of 7:9:9:1. Transfected COS7 cells were identified by binding to anti-CD8 antibody-coated beads (Dynal, Oslo, Norway).19
Forty-eight to 72 hours after transfection, standard whole cell clamp methods were applied to record calcium channel activities.20 The cells were bathed in extracellular solution composed of (in mM) 140 tetraethylammonium Br, 3 KCl, 1 NaHCO3, 1 MgCl2, 1 BaCl2 unless indicated otherwise. The intrapipet solution contained (in mM) 140 Cs-aspartate, 5 egtazic acid, 2 MgCl2, 0.1 CaCl2, 2 K2ATP, 10 N-[2-hydroxyethyl]piperazine-N′-[2-ethanesulfonic acid (HEPES), pH7.4 (Sigma, St. Louis, MO). Ba2+ currents were elicited by step depolarizations from a holding potential of −80 mV using pClamp8 (Axon Instruments, Foster City, CA), with P/5 leak subtraction protocol.
Functional studies of neuromuscular transmission in vivo.
We studied by EMG the three index patients with known EA2 mutations to investigate whether an abnormality in neuromuscular transmission accounts for the patients’ complaint of severe weakness during ataxic spells. For comparison, we examined three additional patients with SCA-6 who have not reported fluctuating weakness. Routine EMG and nerve conduction studies were performed to rule out neuropathy and myopathy, conditions that can also produce abnormal SFEMG findings. Findings were normal.
We used the voluntary-activated SFEMG technique21 to record pairs of action potentials that were time-linked and thus belonging to the same motor unit. The right extensor digitorum communis muscle was used in all subjects. A Nicolet (Madison, WI) SFEMG recording needle electrode was used, and a Nicolet Viking IV instrument and software in which the sweep is triggered by the first spike of the pair, and the variation in the time delay to the second spike of the pair, i.e., the “jitter,” is measured. This time delay is related to the rate of rise of the motor end-plate potential induced by a motor nerve firing and is thus an indirect measure of the process of cholinergic neuromuscular transmission. The mean consecutive difference (MCD) is the mean of differences of the intervals between pairs, calculated by the software, and is a standard parameter for estimating neuromuscular transmission. Blocking of one of the spikes of a pair under SFEMG study occurs when the end-plate potential slope is sufficiently low as to fail to reach threshold. Blocking is another characteristic feature of impaired neuromuscular transmission demonstrated by the SFEMG technique.
Results.
Mutation analysis.
Mutation screening by SSCP and denaturing HPLC revealed an altered elution profile of exon 26 in the index patient in kindred 3 but not in his relatives or 96 normal control subjects (figure 2). Direct sequencing of exon 26 showed a single nucleotide change from T to G at position 4486 of the coding sequence, predicting a change from phenylalanine to cysteine at codon 1406 in the putative P loop close to the membrane-spanning S5 region of domain III of the Cav2.1 calcium channel subunit (figure 3). This mutation was not found in any of the patient’s relatives, including the sister who showed the same haplotype (see figure 1). We detected no expansion of CAG repeats in the patient or his relatives.
Figure 2. Elution profiles of exon 26 by denaturing high-performance liquid chromatography. (A) Detection of heteroduplex molecules in Patient 3. (B) Absence of heteroduplex molecules in the patient’s unaffected sister. The same profile was observed in normal controls.
Figure 3. Diagram of the Cav2.1 subunit with four homologous domains I through IV, each with six transmembrane segments. The locations of episodic ataxia type 2–causing mutations in Patients 1 through 3 are indicated.
Functional studies of heterologously expressed mutant channels in vitro.
Whole cell patch clamp recordings on heterologously expressed mutant channels in COS7 cells demonstrated markedly diminished current density compared with wild type (figure 4; table 1). We detected no endogenous calcium channel activity in untransfected COS7 cells (n = 23) or COS7 cells transfected with plasmids encoding the auxiliary subunits but not the Cav2.1 subunit (n = 21). Of the cells transfected with the wild-type clone, 70% of CD8-positive cells showed Ba2+ current through wild-type calcium channel complexes. Although the same amounts of the wild-type and mutant constructs were used for transfection, a much smaller fraction of CD8-positive cells demonstrated mutant calcium channel activities compared with wild type (see table 1). Furthermore, dramatic reductions in current amplitude were observed compared with wild type (see figure 4 and table 1).
Figure 4. Whole cell patch clamp recordings from transfected COS7 cells with 15 mM Ba2+ in the bath. (A) Ba2+ current elicited with 10-mV step depolarizations from a holding potential of −80 mV. Left, Steps from −20 mV and 20 mV. Right, 10-mV steps from 30 to 80 mV. (B) Current-voltage relationship of wild-type (n = 7; open circle) and 1406C mutant (n = 7; filled circle) calcium channel activities. Vm: membrane potential. Whole cell current amplitude was normalized by cell capacitance. Cells with current amplitude >1.5 nA were excluded.
A comparison of the maximum current density (whole cell current amplitude normalized by cell capacitance) of mutant and wild-type calcium channel complexes expressed in COS7 cells
Functional studies of neuromuscular transmission in vivo.
In SFEMG recordings from normal individuals, the “jitter” in timing of a pair of spikes, as indicated by the MCD measurement, is <50 μS for any single pair, and <35 μS for the average of 20 pairs studied in a muscle. An increased MCD in a maximum of one of 20 recorded pairs is allowed, but neuromuscular blocking is never seen in subjects. In contrast, when neuromuscular transmission is impaired, the low slope of the end-plate potential causes increased variation in the timing and even failure of firing of the muscle fiber action potential, reflected as increased MCD, and blocking in the SFEMG recording. As indicated in table 2, Patients 1, 2, and 3 had abnormally increased mean MCD measurements, more than two individual pairs with increased MCD, and at least one pair with neuromuscular blocking. As a comparison, the respective values for the three patients with SCA-6 were normal, although Patient 6 did have borderline values. An example of a normal SFEMG jitter recording and also an abnormal pair with increased jitter and blocking from Patient 2 is shown in figure 5. Thus, in addition to our findings that the CACNA1A mutations cause abnormalities in calcium currents in vitro, the patients with these mutations have abnormal neuromuscular transmission in vivo.
Results of single-fiber electromyography in subjects with known mutations
Figure 5. Paired recordings by single-fiber electromyography from the right extensor digitorum communis muscle of Patient 2. (A) A normal pair with a normal mean consecutive difference (MCD) of 19 μs. (B) An abnormal pair with increased jittering (MCD = 78 μs) and blocking, demonstrated in (C). Sweep speed was 0.5 ms/division. Gain was 500 μV/division.
Discussion.
We describe three patients with myasthenic syndrome and acetazolamide-responsive EA2. Nonsense mutations in CACNA1A were found in Patients 1 and 2.13,14⇓ In Patient 3, we identified a missense mutation in CACNA1A not present in any other family members. Although no DNA sample was available from the patient’s deceased father to ascertain the absence of this mutation in his genome, the proband’s asymptomatic sister shared the proband’s haplotypes near the CACNA1A gene but did not carry the mutation, strongly suggesting that the mutation arose spontaneously.
Although there appears to be a prevalence of truncation mutations causing EA2 and missense mutations causing FHM,1,12⇓ data from our laboratory and others suggest that this genotype-phenotype correlation is imperfect. We have described a missense mutation G293R causing progressive ataxia with episodic features.23 In addition, unlike most of the other missense mutations in Cav2.1 identified thus far, the F1406C mutation in this report causes EA2 without associated migrainous headaches or other symptoms suggestive of migraine. Indeed, the clinical features associated with F1406C are indistinguishable from those associated with either R1281X or R1549X, two previously reported nonsense mutations causing EA2. Therefore, the type of genetic defect does not predict the clinical profile.
To investigate how these EA2-causing mutations affect calcium channel function, we introduced each of these EA2-causing mutations into a wild-type full-length cDNA clone. R1281X predicts a truncated product containing only the first two domains, and R1549X predicts a product with three of the four domains. F1406C located in the P loop proximal to S5 in domain III could affect ion permeation or disrupt pore formation. We expressed the mutant constructs in COS7 cells, which have no endogenous calcium channel subunits. In untransfected cells as well as those transfected with the auxiliary subunits, we were unable to detect any calcium- or divalent cation-selective current in response to plasma membrane depolarization. Patch clamp recordings of heterologously expressed mutant constructs showed that a missense mutation F1406C and two nonsense mutations R1281X and R1549X causing EA2 produced mutant channel proteins with a dramatic reduction in calcium conductance (see figure 4 and table 1) not observed in SCA-6 or FHM mutations studied to date.
These are loss-of-function mutations because reduced activities or amounts of the mutant calcium channel proteins likely contributed to the marked decreases in calcium currents. At this time we cannot distinguish between haploinsufficiency and a dominant negative effect as the mechanism underlying the loss-of-function mutations that we have identified. A recent report described a complete loss of function of a missense EA2 mutant expressed in HEK293 cells.11 The discrepancy between that observation and our finding of marked but incomplete decrease in mutant calcium channel activity is probably due to differences between COS7 and HEK293 cells as well as differences in the mutations. In particular, there is endogenous calcium channel activity in HEK293 cells, in contrast to the absence in untransfected COS7 cells or those transfected only with the auxiliary subunits.
That we detected calcium channel activities in cells transfected with the nonsense mutants R1281X and R1549X and the missense mutant F1406C suggested that the mutant constructs are transcribed, translated, and targeted to the plasma membrane. These mutant proteins retain the recognition site in the I-II linker for association with the auxiliary β subunit important for plasma membrane targeting. Yet, the mutant proteins may not be folded properly, so that plasma membrane targeting of these mutant proteins can be inefficient.
We detected calcium channel activities less frequently from the R1281X mutant than the R1549X mutant, because only one in 18 R1281X- transfected cells recorded (6%) showed trace Ba2+ current compared with four of 16 R1549X-transfected cells recorded (25%) (table 3). The possibility of read-through, or skipping of the stop codon, cannot be ruled out. It is also possible that the truncated proteins can coassemble to form functional complexes, as suggested by Scott et al.,24 who observed a two-domain short form of Cav2.1 that was associated with the native P/Q-type channels. Perhaps the two-domain R1281X mutant is structurally less complete than the three-domained R1549X mutant and is targeted to the plasma membrane less efficiently.
We describe fluctuating weakness as a prominent complaint in three unrelated patients with EA2. In Patient 3 in this report, myasthenic symptoms preceded the onset of ataxic spells by many years. Patients 1 and 2 had diffuse weakness during ataxic spells. We hypothesized that dysfunctional mutant Cav2.1 calcium channel subunit could interfere with normal calcium entry into the presynaptic nerve terminal, leading to impaired neuromuscular transmission, thus causing weakness.14 Using SFEMG, we demonstrated abnormal neuromuscular transmission in vivo in three patients with CACNA1A mutations, providing strong evidence that the demonstrated mutations are expressed at the neuromuscular junction and that they contribute to the symptoms of fatigue and weakness described by these patients. In contrast, patients with SCA-6 with CAG repeat expansions in CACNA1A had normal neuromuscular transmission. The abnormal neurotransmission at the neuromuscular junction may reflect similarly abnormal neurotransmission at synapses in the CNS in patients with loss-of-function CACNA1A mutations.
Acknowledgments
Supported by UCLA Neurology and NIH grants DC00162, GM43459, and P01 DC02952. C.J.C. received a scholarship from the German National Scholarship Foundation.
- Received May 4, 2001.
- Accepted August 6, 2001.
References
Letters: Rapid online correspondence
REQUIREMENTS
If you are uploading a letter concerning an article:
You must have updated your disclosures within six months: http://submit.neurology.org
Your co-authors must send a completed Publishing Agreement Form to Neurology Staff (not necessary for the lead/corresponding author as the form below will suffice) before you upload your comment.
If you are responding to a comment that was written about an article you originally authored:
You (and co-authors) do not need to fill out forms or check disclosures as author forms are still valid
and apply to letter.
Submission specifications:
- Submissions must be < 200 words with < 5 references. Reference 1 must be the article on which you are commenting.
- Submissions should not have more than 5 authors. (Exception: original author replies can include all original authors of the article)
- Submit only on articles published within 6 months of issue date.
- Do not be redundant. Read any comments already posted on the article prior to submission.
- Submitted comments are subject to editing and editor review prior to posting.
You May Also be Interested in
Hemiplegic Migraine Associated With PRRT2 Variations A Clinical and Genetic Study
Dr. Robert Shapiro and Dr. Amynah Pradhan